The operational and storage stability of tyrosinase biosensors were investigated for different tyrosinase modified electrodes, i.e., plain bulk modified carbon paste electrodes (CPEs), surface modified by simple adsorption to solid graphite electrodes (SGEs), and surface modified by the immobilisation in Eastman AQ, a poly ester–sulphonic acid cation exchanger, and Nafion, a perfluorinated-sulphonated ionomer, on the surface of both CP and SGEs. Factors such as the pH of immobilisation, the enzyme loading, and the polymer concentration were investigated in regards to the influence on the sensitivity, the limit of detection (LOD), the sample throughput (STP), and the operational and storage stability. Both storage and operational stability were improved by immobilisation of tyrosinase in either of the two polymer matrices. For 50 consecutive injections of 50 μM catechol, the response stayed the same for the optimal Eastman and Nafion modified CP and SGEs. After about 42 days 80% and 75% of the original response for the Eastman and Nafion modified tyrosinase electrodes remained, respectively, whereas the tyrosinase bulk-modified CPE and adsorbed tyrosinase SGEs had lost virtually 100% of the original response. The selectivity for nine phenolic compounds were investigated and found to change with the introduction of the polymer membrane. The detection of especially monophenols was improved probably due to their preconcentration into the polymer membranes. The best performing biosensor in terms of sensitivity, LOD, STP, operational and storage stability was reached for one where tyrosinase was immobilised in Nafion, i.e., sensitivity: 11.51 nA/μM, LOD: 0.015 μM catechol, and STP: 36 samples/h. The phenolic content, expressed as catechol equivalents was evaluated in six waste water effluents from tannery industries in Spain and Sweden. The operational stability after 90 consecutive injections of extremely contaminated waste waters showed that the Nafion sensor retained 70% of its initial response.